Ramrao Adik Institute of Technology

Department of Computer Engineering

BE Major Project
On
Swachh A-Eye Garbage Detection System
By
Roll No. Name of Students

21CE1020 Amit Javkar

21CE1021 Aman Sahay
21CE1389 Parth Patil

Mrs. Neha Mahajan

(Major Project Guide)
Outline

• Introduction

• Literature Survey of the existing systems

• Limitations of the existing systems

• Problem statement

• Proposed Methodology

• Design and Architecture

• Hardware and Software Requirements

• Results

• Conclusion

• References

• Publication Details

• Project Competition

2
Introduction

• Waste management is a critical issue in urban and campus environments,

affecting hygiene, sustainability, and public health. Traditional methods rely
on manual inspection, which is inefficient and time-consuming.
• This project aims to develop an AI-powered detection system using
computer vision and IoT technologies to automate cleanliness monitoring
and waste identification.
• The system will use mobile devices to capture images, which will be
processed using deep learning models (YOLOv8) to detect and classify
waste in real-time.
• Implementing this solution can enhance waste management efficiency,
reduce manual effort, and promote a cleaner, more sustainable
environment by enabling timely waste collection and disposal [1].

3 BE Major Project
Literature Survey of existing systems
• Swachhata-MoHUA: It has been created by the Ministry of Housing and Urban
Affairs under the Swachh Bharat Mission. The mobile app enables citizens to report
garbage-related complaints, such as overflowing garbage cans, garbage dumps,
and littered streets, directly to municipal authorities [2].

• Safaimitra Suraksha Challenge: It is aimed to encourage cities to mechanise their

sewer and septic tank cleaning operations in order to prevent deaths and fatalities of
sanitation workers from hazardous entry. Ministry of Housing and Urban Affairs
(MoHUA) launched the Safaimitra Suraksha Challenge (SSC) on the occasion of
World Toilet Day (19th November 2020).

• Recykal: An online waste commerce platform, bridges waste generators

(households, institutions, and organisations) with recyclers and waste management
firms.

• Aakri: A startup based in Hyderabad that offers a platform for scrap collection
services, allowing households and small organisations to eliminate dry waste such
as paper, plastic and metal.

4 BE Major Project
Literature Survey of existing systems

5 BE Major Project
Literature Survey of existing systems
• Bin-e : A European-designed intelligent waste bin solution that uses AI-powered
object recognition and ultrasonic sensors to automatically sort waste into the
appropriate compartments.

• Evreka : An integrated smart waste management system that has a range of

products such as smart bin sensors, mobile workforce apps and route optimisation
for collection vehicles. It uses real-time information from sensor-enabled bins to
dynamically adjust collection schedules.

6 BE Major Project
Limitations of the existing systems

• Many AI-based garbage detection systems struggle with varying

environmental conditions such as poor lighting reducing detection accuracy.
• Training AI models requires large datasets with labeled waste images, which
can be time-consuming and resource-intensive to collect.
• Deep learning models require significant processing power, making it difficult
to deploy them on low-cost edge devices for real-time waste detection [3].
• Implementing AI and IoT-based waste management solutions at a large scale
can be expensive, limiting adoption in resource-constrained areas like small
campuses or underfunded municipalities.
• Identifying and classifying different types of waste (biodegradable,
non-biodegradable, hazardous) accurately remains a challenge, affecting
proper waste disposal and recycling.
• Municipal waste management systems may not be equipped to integrate
AI-based detection tools seamlessly, requiring additional investments in
hardware and software [4].

7 BE Major Project
Problem Statement

Maintaining cleanliness is challenging due to scattered waste, lack of real-time

monitoring, and inefficient manual inspection. Traditional methods rely on periodic
checks, which may lead to delays in waste removal and an unhygienic
environment. A Garbage Detection system aims to address this issue by using AI,
computer vision, and IoT sensors to automatically detect littered areas and monitor
cleanliness levels. This system ensures a cleaner, healthier, and more sustainable
campus environment by enabling proactive waste management and efficient
resource allocation.

8 BE Major Project
 Proposed Methodology

1. Data Collection: Capture live video feeds from multiple cameras and gather
sensor readings from the smart bin.

2. Garbage Detection: Use the YOLOv8 model to detect garbage objects in

real-time from video frames.

3. Sensor Monitoring: Continuously read dustbin fill levels using ultrasonic sensors
connected to Arduino.

4. Status Update: Update room and bin status based on detection results and
sensor data.

5. User Interface: Display live video, bin status, and room cleanliness status through
a GUI built with Tkinter.

9 BE Major Project
Smart Dustbin
The smart bin is equipped with an ultrasonic sensor connected to an Arduino. The
ultrasonic sensor continuously measures the distance from the sensor to the trash level
inside the bin. Based on the measured distance, the Arduino determines whether the bin
is empty, half-filled, almost full, or full. These status updates are displayed in real-time on
a graphical user interface (GUI) built using Tkinter in Python [6].

Features & Functionalities:

1. Real-time Fill Level Monitoring – Uses IoT sensors (ultrasonic) to detect how full
the bin is.
2. Automated Notifications – Sends alerts when the bin reaches a predefined
threshold (e.g., 80% full).
3. Priority-Based Cleaning System – Full bins are marked as high priority, ensuring
timely emptying and optimized resource allocation.

10 BE Major Project
Smart Dustbin

11
System Design

12 BE Major Project
Hardware and Software Requirements
Hardware Requirements

A computer with enough processing capacity (a multi-core processor, 8GB RAM, and a
dedicated GPU for real-time object detection), a webcam, and an Arduino with an ultrasonic
sensor for measuring distance to detect the fill level of the trash cans. The ultrasonic sensor
(HC-SR04) is interfaced to the Arduino and continuously feeds the system with real-time
distance information. The Arduino is connected to the computer through a serial USB
interface (COM port). A stable internet connection might also be required for downloading
the necessary software libraries and models.

Software Requirements

The important software components include Python as the primary programming language,
OpenCV for video processing and managing the camera inputs, and YOLOv8 (pre-trained
deep learning model) for object detection. Tkinter, along with the ttkthemes library, is used
to build the graphical user interface (GUI) to display real-time data and video feeds. The
system also uses the Ultralytics library for YOLO model integration. The Arduino
communicates with the Python program via the PySerial library for real-time sensor data
transfer. It is designed to work on Windows OS based Python 3.x and utilises pip-installed
necessary libraries.

13 BE Major Project
Results

Monitoring Dashboard Real time waste classification

14 BE Major Project
Results

15
Conclusion

Despite the challenges, the system proves to be a cost-effective and scalable

solution for waste detection and management. The AI model’s efficiency, combined
with IoT-based smart bins, provides an intelligent approach to waste collection and
monitoring. The system significantly reduces the dependency on manual waste
detection and collection methods, leading to more streamlined and sustainable
waste management practices.

16 BE Major Project
References

[1] Ferronato, N., & Torretta, V. (2019). "Waste Mismanagement in Developing Countries:
A Review of Global Issues." Waste Management, 87, 139-151.

[2] Swachh Bharat Mission (Government of India). "Technology Interventions in Waste

Management." Ministry of Housing and Urban Affairs Report, 2022.

[3] Singh, R., & Gautam, S. (2021). "AI-Enabled Smart Waste Management Systems: A
Comparative Study." Journal of Cleaner Production, 295, 126450.

[4] Gupta, S., & Sharma, A. (2019). "A Machine Learning Approach to Garbage
Classification for Smart Cities." IEEE Transactions on Sustainable Computing, 4(3),
476-487.

[5] Jin, Y., Zhao, C., Li, W., & Bai, J. (2021). "Automatic Garbage Detection in Urban
Environments Using Deep Neural Networks." Sustainability, 13(5), 2895.

[6] Saha, M., Paul, S., & Kumar, N. (2020). "Smart Waste Management Using IoT-Based
Sensors,Drones, and AI." International Journal of Environmental Science and
Technology, 17(9), 2345-2358.

17 BE Major Project
Publication Details

1. N. Mahajan, A. Javkar, A. Sahay, P. Patil, 'Swachh A-Eye Garbage Detection

System', Sixth World Conference on Artificial Intelligence: Advances and
Applications (WCAIAA), May 2025. [Selected for presentation]

18
Project Competition

1. Microsoft AI Agents Hackathon 2025

19
Thank You